An Engine denotes a Product Insights AI Model which has been run over the products and messages present in the workspace/ dashboard. The type of AI-enrichment that happens over a message is determined by what engine has been run over the message. There can be multiple engines deployed on a workspace.

Product Insights – Engine (Product) is the dimension used to filter out engine-specific mentions. This filter can be applied on Dashboard, Section, and Widget levels.

Inside Product Configuration, the Vertical section contains the name of the model. The engine corresponding to the model will be run on that product’s mentions.

No – only one engine can be run on a specific message

The engine and the vertical/ model have a slight difference in their naming conventions.

Suppose the PI model’s name is "Television", then the corresponding engine name will be "piTelevision". These two are separately named and the naming can be varied too. But a PI engine’s name will always have the prefix of "pi".

For live data, the engine corresponding to the Vertical name from the products gets run on those sets of products. To run the engine on historic data, you can contact the Product Team.

The Engine B will overwrite Engine A, and the PI enrichment will now be corresponding to Engine B, not Engine A.